$585B on new fabs, $357B on AI data centers, and $5.8B on humanoid robots. One of those numbers is not like the others
We also don't yet know how to be as efficient with training examples as any living creatures' brain, and we only partially make up for this by training on so many examples it would take you a million or so years to do the same, so we'd still stuggle with something proportionally smaller-brained such as a cat.
That said, remote controlled androids are going to be economically disruptive, as they make every (unlicensed) job open to outsourcing from an office in a low wage country.
We don't know the violations of the physical Church-Turing thesis that are conductive for machine learning. We don't have evidence for their existence in the brain (although, the brain would be the prime candidate for finding them as evolution works directly with the true physical laws).
BTW, large ANNs don't try to model how the brain does things. They are trying to mimic what the brain does. So, using "how many transistors/artificial neurons it takes to model a biological neuron" is not a good approach.
We have no evidence. We even have no solid theories how this can work (Penrose's OrchOR is "OrchOR somehow taps into mathematical knowledge somehow encoded into the structure of spacetime"). But people, for some reason, insist that there should be something there. I can't attribute it to anything else but to deeply entrenched feeling of human exceptionalism.
Look up the neural correlates replication crisis, and e. g. the "dead salmon" study by Bennett et al.
Self driving cars is far from a solved problem. It's something we've been working on for the last 20 years at least. We are getting closer, some solutions are impressive (like waymo). But even those ultimately need operators in the area of the cars to solve the problem of the car getting stuck.
Self driving cars are an infinitely simpler problem to solve vs a general purpose humanoid robot. You have basically 2 outputs, acceleration and steering. You have rules simple enough that I was driving at 14. With enough input, you'd think self driving could be completed in a snap. But it's not there yet.
Humanoid robots which are useful will come after widespread deployment of L5 self driving cars. Since we don't have that, I have no faith that we are close to useful humanoids.
"I have no faith that we are close to useful humanoids." I agree completely. We still have competitions about having robots walk up stairs and open doors, and balance while doing those things...
What would be the top uses of humanoid robots?
To replace humans for general tasks. McDonalds would like to use them to flip burgers, take orders, and deliver meals. Amazon would like to use them to pick items from their warehouses for delivery.
I'd argue, though, in both cases the "humanoid" part isn't what they really want or need. Even in the mcdonalds case, so long as the robot is "cute" enough to interact with customers, then it can be a box with an arm.
Of course, a major problem with using robots in the food industry is cleaning and hygiene. Robots tend to have lots of hard to clean cracks that will happily carry around germs. Even in medicine, we are finding that things like laparoscopic devices are beasts that are almost impossible to fully clean.
* Tasks where poorly-paid humans are cheaper than expensive factory robots. Humanoid robots are more complex / fiddly so will be more expensive than existing factory robots. No help here.
* Tasks where human dexterity has an advantage over state-of-the-art robot actuators (e.g. sewing fabric panels into garments). Better robotics could help here, but the advancement needed is better actuators, not AI and a humanoid form-factor. And if you solve this you'd be better off putting your new end-effector on an existing 6dof platform.
* Supervising robotic equipment and handling exceptions. But then you get to "handling poorly-specified unfamiliar tasks in the physical world" which is not currently a solved problem and there's no guarantee just throwing more compute at it will be sufficient to solve this. So far all humanoid robot demos have either been either known tasks in a tightly-controlled environment, teleportation, or so poorly-functioning as to be obviously not fit for purpose.
I'd also point out that what makes this whole thing smell of being a grift is the fact that what's being chased is humanoid.
Humans are not the pinnacle of dexterity or stability. And freed from biological constraints, it makes no sense why we'd use the human form as a reference.
Making a humanoid robot is a hard thing to do, for sure, but it's also not particularly useful. The routines needed to balance a robot or correct for a slip are interesting to solve, but don't really make for a better robot which is more capable of doing the dishes or folding the laundry.
If I'm amazon, for example, then the most useful form factor for a general purpose robot is 2 arms on a 4 wheel omnidirectional rolling platform.
but personally, I would welcome our cyper-coon overlords.
An artificial manipulator with even half of these features would be the holy grail of robotics. Doing them all at once is a miracle.
That's what I'm calling a grift of the humanoid robots. They are focusing pretty heavily on making manipulators shaped like hands, which don't perform anywhere near hand capabilities. The form is mattering more than the functionality which introduces a lot of unnecessarily hard to solve functionality problems.
The actual hard part of doing something like folding cloths isn't even the manipulation (though it's part of it) you mostly just need a few pincers. The hard part is that fabric changes like crazy in a 3 dimensional way on every motion. Just identifying "this is a shirt" is a hard problem to solve. Further deciding how to fold said shirt is extra difficult. But even further, getting a robot to know "this is where this shirt should be put away" and "this is how I should handle new cloths" is crazy hard. That's the part that I've seen absolutely no evidence that any of these humanoid robots are making any sort of progress on. The most impressive ones are cheating with a guy in the room over doing the task. Not exactly something I'd want in my home. Perhaps in a nuclear reactor.
I've highlighted the two main issues you are currently experiencing.
They're being conservative with their rollout mostly because of civic issues. Most places do not have a legal framework for "what if your autonomous vehicle hits someone?" yet. Even if Waymos never were at fault for a collision with a person, you can always have cases like the one in Georgia back in October where a bicyclist wasn't looking where they were going and rammed into one. The shaky legal ground is a pretty big impediment right now, and that's in the US where we have much laxer laws about corporations killing people.
Then, there will be concerns about the scaling costs. If it makes economical sense only in densely populated zones, what's the point? I mean, yes, it will be yet another business that works only in cities in that case, but that will not then make it universal as cars are.
More basic movement control doesn't need loads of ram as far as I know.
The larger the context window, the better with models. Having a few TB of RAM would be exceptionally helpful.
All this just made me realise something however. Having your robot dormant and charging, is a bit of a waste. You could have robots dormant, but its compute in use to act as a compute node. If the distribution of robots is similar world-wide, we'd need a fraction of the datacenters we have now.
Using such nodes for training purposes would be beyond advantageous. And the company which can slice up the work and having training done in batches would get the big bucks. And actually, with consumer facing products soon all laden with extra ram and gpu for local compute, that applies there too.
Imagine leasing out idle time on your desktop or even laptop for cash. There may be a market here, especially with the cost of new datacenters. Any company able to securely package compute without risking data safety is going to make a mint.
Anyone have any ideas?
I don't think you understood my post. The equivalent of self-driving is the movement control I was talking about.
Self-driving cars don't have high level logic, except for route planning. Which often is offloaded to the cloud. An extra 30 milliseconds on understanding your speech is nothing.
> Imagine leasing out idle time on your desktop or even laptop for cash.
The same is true with an android. Imagine it turning on an frying pan, cooking dinner, and then going offline part way through. Or turning on a tap to wash something, and going offline while the sink overflows and destroys the house.
There are myriad of such scenarios, but local compute is absolutely, 100% necessary. Anyone betting the farm on putting network controlled devices into homes for any serious task is going to lose their shirt. Local compute is an absolute requirement, and a few TB of RAM and local compute will be nothing over the scope of this discussion (a few years minimum, just to build and kick off all these new fabs).
By the time these fabs are online, expect most smart phones to have 1TB of RAM and significant llm capable compute (gpu or other custom silicon). I would be astonished if flagship model phones in 2030 weren't sold with 1TB RAM. Note I'm saying flagship, there will be of course economy models as always. Certainly laptops will be sold in multi-TB RAM configs.
You don't need terabytes to turn things back off.
Not that I want to trust "not setting the house on fire" and "not flooding the house" to this kind of model in the first place...
> I would be astonished if flagship model phones in 2030 weren't sold with 1TB RAM.
I'll be astonished if they have 50GB.
Have you looked at RAM size/price trends? I'm not even talking about the last year, just the pattern before that. We're not in the 80s and 90s anymore. The most recent price lows were roughly $3.50/GB in 2013, $2.50/GB in 2016, and $1.50/GB in 2023. If we're lucky the cheapest stuff will hit $1/GB in a few years, and the kind that would actually fit on a phone motherboard would be significantly more than that.
Samsung hit 16GB on their top model in 2020 and it's either been 16GB or 12GB ever since. Apple only went up to 12GB in the last year. Google offers 16GB. A couple niche offerings have 24GB. Why would these RAM numbers even double during the next four years?
Here's the thing, by 2028, maybe 2029, 1/2 the cloud AI providers are going to go bankrupt, or severely downsize. This will likely be due to massive reduced power and processor requirements for AI. GPUs are a stopgap, and they won't be used much longer. When that happens, and the number of servers drops by 90%, along with similar power requirements, a lot of people are going to go tits up due to massive, empty datacentres and compute.
That will absolutely flood the market with cheap obsolete GPU + RAM, used servers, you name it. Current RAM prices will plummet, and those able to withstand the storm will have to compete again, not just ride the wave like they've been doing now.
RAM has been ridiculously high for a long time, it won't be by 2030.
Couple that with local SoC with dedicated compute for LLMs, and you have the perfect storm for local, on phone compute. Why this? Well, because LLMs will be in everything! Certainly in robots, autonomous weapons, and anything the crazies can shove it into. Phones won't be the unique case, it'll be sharing a hardware market with robots, and 100 other things using the same SoCs.
Local compute will be easy... RAM makes a massive difference, huge as you know, difference in local compute. And with the datacentre market in tatters, and people becoming more and more concerned about privacy, people will likely not want to whisper to a cloud LLM, about their fetishes for donkeys or moose or whatever turns their crank.
Not to mention, other aspects of their lives.
In as RAM will be super cheap, that's where I see this going.
I have a sneaking suspicion you don't agree with this assessment. I guess we'll come back here 2030 and see what's what.
Also "massively reduced requirements for AI, so much so that we have 90% fewer servers despite the Jevons paradox" is not compatible with "high quality models make effective use of terabytes of memory" and "LLMs will be in everything".
And even if you could perfectly reuse all the memory out of the collapsing datacenters, that's not enough to make many of the phones and laptops you're describing. We'd need a 10x increase in manufacturing on top of the massive price drops, and even if we started building those fabs today they wouldn't be done in time (and we wouldn't have enough equipment to fill them).
I see... tech giants and R & D at the top of the pyramid, and I have seen Gumee, the fab that started all this Korean ram prodution.